def main():
config = get_config_from_json('config.json')
# create an instance of the model
model = DEC(config)
# create trainer instance
trainer = Trainer(config, model)
# train the model
trainer.train()
def main():
config = get_config_from_json('config.json')
# create an instance of the model
model = VAE(config)
# create experiments instance
experiments = Experiments(config, model)
# create trainer instance
trainer = Trainer(config, model, experiments)
# train the model
trainer.train()
utils.save_dict_to_json(params, json_path)
# Launch training with this config
cmd = "{python} train.py --model_dir {model_dir} --data_dir {data_dir}".format(
python=PYTHON, model_dir=model_dir, data_dir=data_dir)
print(cmd)
subprocess.check_call(cmd, shell=True)
if __name__ == "__main__":
# Load the "reference" parameters from parent_dir json file
args = parser.parse_args()
json_path = os.path.join(args.parent_dir, 'params.json')
assert os.path.isfile(
json_path), "No json configuration file found at {}".format(json_path)
params = utils.get_config_from_json(json_path)
# Perform hypersearch over one parameter
# image_size = [128, 224, 512, 1024]
# batch_size = [64, 32, 16, 8]
image_size = [1024]
batch_size = [4]
for sz, bz in zip(image_size, batch_size):
# Modify the relevant parameter in params
params.image_size = sz
params.batch_size = bz
# Launch job (name has to be unique)
job_name = "image_size_{}_bz{}".format(sz, bz)
launch_training_job(args.parent_dir, args.data_dir, job_name, params)